From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 18220 invoked by alias); 9 Oct 2011 11:04:32 -0000 Received: (qmail 18209 invoked by uid 22791); 9 Oct 2011 11:04:32 -0000 X-SWARE-Spam-Status: No, hits=-2.9 required=5.0 tests=ALL_TRUSTED,AWL,BAYES_00 X-Spam-Check-By: sourceware.org Received: from localhost (HELO gcc.gnu.org) (127.0.0.1) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Sun, 09 Oct 2011 11:04:18 +0000 From: "Joost.VandeVondele at pci dot uzh.ch" To: gcc-bugs@gcc.gnu.org Subject: [Bug libfortran/50673] New: very slow I/O with trailing spaces Date: Sun, 09 Oct 2011 11:04:00 -0000 X-Bugzilla-Reason: CC X-Bugzilla-Type: new X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: gcc X-Bugzilla-Component: libfortran X-Bugzilla-Keywords: X-Bugzilla-Severity: normal X-Bugzilla-Who: Joost.VandeVondele at pci dot uzh.ch X-Bugzilla-Status: UNCONFIRMED X-Bugzilla-Priority: P3 X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org X-Bugzilla-Target-Milestone: --- X-Bugzilla-Changed-Fields: Message-ID: X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/ Auto-Submitted: auto-generated Content-Type: text/plain; charset="UTF-8" MIME-Version: 1.0 Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Archive: List-Post: List-Help: Sender: gcc-bugs-owner@gcc.gnu.org X-SW-Source: 2011-10/txt/msg00631.txt.bz2 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50673 Bug #: 50673 Summary: very slow I/O with trailing spaces Classification: Unclassified Product: gcc Version: 4.7.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: libfortran AssignedTo: unassigned@gcc.gnu.org ReportedBy: Joost.VandeVondele@pci.uzh.ch The following testcase (derived from CP2K, which required 20min to read a 600Mb file) is about 100x times slower with gfortran than with ifort (12.0.4) CHARACTER(LEN=40480) :: line="O 0.12456789 0.123456789 0.123456789" CHARACTER(LEN=2) :: AA REAL*8 :: vec(3) DO i=1,10000 read(line,*) AA,vec ENDDO END The issue seems related to how efficient the trailing spaces are handled in both compilers. 4.7 is a bit (20%) slower than 4.3, but nothing fundamental. Profiling the code shows that most time is spent in next_char, mem_read, memcpy, eat_spaces.